JP6335676B2 - Substrate type optical waveguide device - Google Patents

Description

  The present invention relates to a substrate type optical waveguide device.

  Currently, the amount of information transmitted by optical communication is steadily increasing. In order to cope with such an increase in the amount of information, measures such as increasing the signal speed and increasing the number of channels by wavelength division multiplexing are being promoted. In particular, in the next-generation 100 Gbps digital coherent transmission technology for the purpose of high-speed information communication, a polarization multiplexing scheme that places information in two polarization modes in which electric fields are orthogonal is used. In this polarization multiplexing system, the amount of information per unit time can be doubled compared to an optical transmission system using a single polarization.

  However, the optical modulation system for high-speed communication including the polarization multiplexing system requires a complicated optical modulator, which causes problems such as an increase in size and cost of the apparatus. In order to solve such a problem, research on an optical modulator using a substrate type optical waveguide element has been conducted.

As an example of the substrate-type optical waveguide element, there is one including a waveguide having a core made of silicon (Si) and a clad made of quartz (SiO ₂ ) having a refractive index smaller than that of the core on the substrate. Such a substrate type optical waveguide device using Si has advantages such as miniaturization by integration and cost reduction by mass production by using Si, which is easy to process and has a high refractive index, as a material.

  Incidentally, an optical modulator including a polarization multiplexing system using such a substrate type optical waveguide device has the following problems. That is, in the substrate-type optical waveguide device, the cross-sectional shape of the core forming the waveguide is asymmetric in the direction parallel to the substrate (width direction) and the direction perpendicular to the substrate (thickness direction). It is common. Therefore, a polarization mode (TE mode) in which the main component of the electric field is the in-plane direction of the substrate and a polarization mode in which the main component of the magnetic field is the in-plane direction of the substrate (the main component of the electric field is the vertical direction of the substrate). Characteristics such as the effective refractive index differ from those of the wave mode (referred to as TM mode).

Of these two polarization modes, the TE ₀ mode and the TM ₀ mode are often used. Among these, the TE ₀ mode is the mode having the largest effective refractive index among the TE modes, and the TM ₀ mode is the mode having the largest effective refractive index among the TM modes.

One of the basic components of an optical device is a branching waveguide (multiplexing waveguide or demultiplexing waveguide) that combines or demultiplexes light. Among them, the Y branch waveguide is generally known. Here, the Y branch waveguide shown in FIGS. 35A, 35B, and 35C will be described. FIG. 35A is a plan view showing a Y-branch waveguide. FIG. 35B is a cross-sectional view of the Y branch waveguide by Z ₇ -Z ₇ shown in FIG. FIG. 35C is an enlarged plan view showing the encircling portion C shown in FIG.

35 (A) and 35 (B) includes two input waveguides 201 and 202, and one output waveguide 203 in which the other end is connected to one end of these two input waveguides 201 and 202. And a clad 205 that covers the core 204 and has a refractive index lower than that of the core 204. The core 204 forms waveguides 201, 202, and 203 having a rectangular cross section made of Si (so-called rectangular waveguides) 201, 202, and 203 on the surface of the lower clad 205. Cladding 205 includes a lower clad 206 made of SiO ₂ or the like, and an upper clad 207 formed of an air layer or SiO ₂ or the like. The upper clad 207 covers the surface of the lower clad 206 on which the waveguides 201, 202, and 203 are formed.

  In the Y branch waveguide shown in FIGS. 35A and 35B, the two parallel input waveguides 201 and 202 are gradually approached toward the side (one end side) connected to the output waveguide 203, In principle, low loss multiplexing is possible by connecting to the output waveguide 203 continuously.

  However, when such a Y-branch waveguide is manufactured, in a portion where the two input waveguides 201 and 202 are connected to the output waveguide 203, ideally, the enclosed portion C shown in FIG. Although it is necessary to connect the two input waveguides 201 and 202 at an acute angle, it is difficult to manufacture such a branched waveguide structure with high accuracy in an actual manufacturing process. In other words, in the actual branching waveguide structure, as shown in FIG. 35C, the two input waveguides 201 and 202 are connected in front of an acute angle. For this reason, the loss in the part which connects the two input waveguides 201 and 202 to the output waveguide 203 will increase.

On the other hand, a Y-branch waveguide described in Non-Patent Document 1 below can be given as a branch waveguide structure that is easy to manufacture and has a relatively low loss. Here, the Y-branch waveguide described in Non-Patent Document 1 below will be described with reference to FIGS. 36 (A), (B), and (C). FIG. 36A is a plan view showing a Y-branch waveguide described in Non-Patent Document 1 below. FIG. 36B is a cross-sectional view of the Y branch waveguide by Z ₈ -Z ₈ shown in FIG. FIG. 36C is a cross-sectional view of the Y branch waveguide by Z ₉ -Z ₉ shown in FIG.

36 (A), (B), and (C), the Y branch waveguide has two input waveguides 301 and 302, and one of the two input waveguides 301 and 302 connected to the other end. A core 304 that forms the output waveguide 303 and a clad 305 that covers the core 304 and has a lower refractive index than the core 304 are provided. The core 304, on the plane of the lower clad 305, for example to form a rectangular cross section of each waveguide (the so-called rectangular waveguides) 301, 302, and 303 made of SiO _2. Cladding 305 includes a lower clad 306 made of SiO _2, and an upper clad 307 formed of an air layer or SiO _2. The upper clad 307 covers the surface of the lower clad 306 in which the respective waveguides 301, 302, and 303 are formed.

  In the Y-branch waveguide shown in FIGS. 36A, 36B, and 36C, the two parallel input waveguides 301 and 302 are not parallel to the output waveguide 303 while being kept parallel to each other. Connected continuously. Further, the output waveguide 303 has two input waveguides 301, 302 so that the electric field distribution immediately before being connected to the output waveguide 303 of the input waveguides 301, 302 is efficiently coupled to the electric field distribution of the output waveguide 303. It has a wide shape protruding from both sides in the width direction from 302. That is, the width of the output waveguide 303 is larger than the sum of the width of the two input waveguides 301 and 302 and the distance between the two input waveguides 301 and 302.

Yuji Matsuura et al., “Low-loss Y-branch device with new structure”, 1994 IEICE Spring Conference, C-330, April 1994

By the way, in the Y-branch waveguide described in Non-Patent Document 1 described above, in order to efficiently couple TE ₀ mode light having an electric field distribution concentrated at the center of the output waveguide 303, one input waveguide is used. It is necessary that the TE ₀ mode light input to 301 oozes into the other adjacent input waveguide 302 and has an electric field distribution that approaches the center of these two input waveguides 301 and 302.

However, in the Y-branch waveguide described in Non-Patent Document 1, since the input waveguides 301 and 302 and the output waveguide 303 are configured by the rectangular waveguide composed of the core 304 having the rectangular cross section described above, the core 304 Light confinement in the width direction of the light is strong, and light penetration into the adjacent waveguide is weak. In this case, since the overlap of the electric field distribution between the two input waveguides 301 and 302 and the output waveguide 301 becomes weak, there is a problem that it is difficult to multiplex with low loss. In particular, the substrate type optical waveguide element has a high refractive index difference between the core made of Si and the clad made of SiO ₂ (including the air layer), so that the light is strongly confined in the core. The problem is remarkable.

  One aspect of the present invention has been proposed in view of such conventional circumstances, and is a substrate-type optical waveguide element capable of multiplexing or demultiplexing low-loss light, and such An object is to provide an optical modulator using a substrate type optical waveguide element, a monitor structure of the optical modulator, and an output polarization conversion element.

  In order to achieve the above object, a substrate-type optical waveguide device according to one aspect of the present invention includes a substrate and a core that covers the core and has a refractive index smaller than that of the core on the substrate. A first optical waveguide element and a second waveguide parallel to each other, and a third optical waveguide element having the other end connected to one end of the first waveguide and the second waveguide. A branched waveguide structure having a plurality of waveguides. The core includes a first rib portion having a rectangular cross section forming the first waveguide, a second rib portion having a rectangular cross section forming the second waveguide, and the third waveguide. A third rib portion having a rectangular cross-section forming the first rib portion, the second rib portion, and the second rib portion at a height lower than the first rib portion and the second rib portion. And a slab portion shared between the three rib portions. In the first waveguide and the second waveguide, the widths of the first rib portion and the second rib portion are constant in the respective length directions, and the first rib portion and the second waveguide portion are A linear waveguide is formed in which the distance between the second rib portion is constant in the length direction of the first rib portion and the second rib portion. In the third waveguide, the center in the width direction of the third rib portion coincides with the center in the width direction between the first rib portion and the second rib portion, and the third waveguide The width of the third rib portion is greater than the sum of the width of the first rib portion, the width of the second rib portion, and the distance between the first rib portion and the second rib portion. A large linear waveguide is formed.

  In the substrate-type optical waveguide element, the slab portion may be provided continuously on both sides in the width direction of the first rib portion and the second rib portion.

  In the substrate-type optical waveguide element, the slab portion may be provided continuously on both sides in the width direction of the third rib portion.

  The substrate-type optical waveguide element has a bent waveguide having one end connected to the other end of either the first waveguide or the second waveguide, or both of the bent waveguide. The first rib portion and the second rib portion are bent from the other end side toward the one end side by bending either or both of the first rib portion and the second rib portion in a plane. The structure which has a shape where the space | interval between rib parts becomes small continuously may be sufficient.

  The substrate-type optical waveguide element has a tapered waveguide having one end connected to the other end of the bending waveguide, and the tapered waveguide is formed on one side or both sides in the width direction of the first rib portion. A first slab portion provided continuously, and a second slab portion provided continuously on one side or both sides in the width direction of the second rib portion, and the first slab portion, The second slab portion may be configured to be continuously provided in the slab portion and have a shape in which each width continuously increases toward the slab portion.

  The substrate-type optical waveguide element has a fourth waveguide having the other end connected to one end of the third waveguide, and the core has a rectangular cross section that forms the fourth waveguide And the fourth waveguide has a taper in which the width of the third rib portion gradually decreases toward the side opposite to the side connected to the third waveguide. The structure which forms the shape-like waveguide may be sufficient.

In the substrate-type optical waveguide element, the core may include Si and the cladding may include SiO ₂ .

  An optical modulator according to an aspect of the present invention includes an input unit to which light is input, a demultiplexing unit that demultiplexes light input from the input unit, and is demultiplexed by the demultiplexing unit At least one phase modulation unit for phase-modulating light; a multiplexing unit for multiplexing the light phase-modulated by the phase modulation unit; and an output unit for outputting the light combined by the multiplexing unit; And any one of the substrate-type optical waveguide elements is used for the multiplexing section.

  An optical modulator according to one aspect of the present invention includes a first input unit to which light is input, a first demultiplexing unit that demultiplexes light input from the first input unit, At least one phase modulation unit that phase-modulates the light demultiplexed by the first demultiplexing unit, a first multiplexing unit that multiplexes the light phase-modulated by the phase modulation unit, and A first output unit that outputs light combined by the first combining unit, a first light modulation unit, a second light modulation unit, and the first light modulation unit and the first light modulation unit, respectively. A second input unit to which light is input and the light input from the second input unit on the first input unit side of the first light modulation unit And a second demultiplexing unit for demultiplexing to the first input unit side of the second light modulation unit, and a stage subsequent to the first light modulation unit and the second light modulation unit, The first light change A second combining unit that combines the light output from the first output unit of the unit and the light output from the first output unit of the second light modulation unit, and the second combining unit. A second output unit that outputs light combined by the wave unit, and any one of the first combining unit and the second combining unit includes any one of the substrate-type optical waveguide elements It is characterized by using.

  The optical modulator adjusts a phase difference of the light output from the first multiplexing unit of either the first optical modulation unit or the second optical modulation unit. May be provided.

The monitor structure for an optical modulator according to one aspect of the present invention is the optical modulator having any of the above optical modulators, wherein the effective refractive index is the largest in the TE mode in which the electric field is in the in-plane direction of the substrate. When the TE ₀ mode light representing the mode and the TE ₁ mode light having the second highest effective refractive index are simultaneously output from the output unit, the TE ₁ mode light is separated and detected. a monitoring structure of the optical modulator, the located downstream of the output section, and the high-order mode splitter for separating the light of the TE ₁ mode, the light of the above TE ₁ mode that has been separated in the higher mode splitter And a photodetector for detecting.

The monitor structure for an optical modulator according to one aspect of the present invention is the optical modulator having any of the above optical modulators, wherein the effective refractive index is the largest in the TE mode in which the electric field is in the in-plane direction of the substrate. When the TE ₀ mode light representing the mode and the TE ₁ mode light having the second highest effective refractive index are simultaneously output from the output unit, the TE ₁ mode light is converted into the electric field by the substrate. This is a monitor structure of an optical modulator that separates and detects the TE ₀ mode light after it is converted to TE ₀ mode light, which represents the mode with the largest effective refractive index in the TM mode in the vertical direction. The high-order polarization converter that converts the TE _1- mode light into the TE _0- mode light, the polarization beam splitter that separates the TE _0- mode light, and the polarization beam splitter. of the TE ₀ mode Characterized in that it comprises a photodetector for detecting a.

An output polarization conversion element according to one aspect of the present invention includes an input unit to which light is input, a demultiplexing unit that demultiplexes light input from the input unit, and demultiplexing by the demultiplexing unit A Mach-Zehnder interferometer having a multiplexing unit that combines the combined light and an output unit that outputs the light combined by the multiplexing unit, and a TE mode in which the electric field is in the in-plane direction of the substrate. The TE ₀ mode light representing the mode with the largest effective refractive index and the TE ₁ mode light having the second highest effective refractive index among the TE modes in which the electric field is in the in-plane direction of the substrate. A high-order polarization converter that converts light into TE ₀ mode, which represents a mode having the largest effective refractive index among TE modes in which the electric field is in the in-plane direction of the substrate, and is demultiplexed by the demultiplexing unit A phase adjustment unit that adjusts the phase difference of one of the received light, and Characterized by using one of the planar optical waveguide device.

  The output polarization conversion element may include two phase modulation units that phase-modulate the light demultiplexed by the demultiplexing unit.

  As described above, according to one aspect of the present invention, a substrate-type optical waveguide device capable of multiplexing or demultiplexing low-loss light, and light using such a substrate-type optical waveguide device It is possible to provide a modulator, an optical modulator monitor structure, and an output polarization conversion element.

The substrate type optical waveguide device concerning a 1st embodiment of the present invention is shown, (A) is the top view, (B) is the sectional view in the input waveguide, (C) is the sectional view in the output waveguide. is there. Is a graph obtained by simulation of an electric field distribution of the even mode (E _X component) when the input waveguides adjacent apart, (A) is a graph showing the electric field distribution, (B) is its Y = 0 is a graph showing the _{E X} component at the position of [[mu] m]. Is a graph obtained electric field distribution of the odd mode (E _X component) by a simulation when the input waveguides adjacent apart, (A) is a graph showing the electric field distribution, (B) is its Y = 0 is a graph showing the _{E X} component at the position of [[mu] m]. It is sectional drawing which shows an example of the dimension of each part of the Y branch waveguide used by the simulation shown in FIG.2 and FIG.3. Is a graph obtained by simulation of an electric field distribution of the even mode (E _X component) when the input waveguide adjacent is approaching, (A) is a graph showing the electric field distribution, (B) is its Y = is a graph showing the _{E X} component at the position of 0 [μm]. Is a graph obtained electric field distribution of the odd mode (E _X component) by a simulation when the input waveguide adjacent is approaching, (A) is a graph showing the electric field distribution, (B) is its Y = is a graph showing the _{E X} component at the position of 0 [μm]. It is sectional drawing which shows an example of the dimension of each part of the Y branch waveguide used by the simulation shown in FIG.5 and FIG.6. Is a graph obtained by simulation of an electric field distribution by the TE ₀ mode of a rectangular waveguide (E _X component), the position of (A) is a graph showing the electric field distribution, (B) is its Y = 0 [μm] it is a graph showing the E _X component at. Is a graph obtained by simulation of an electric field distribution due to TE ₁ mode of the rectangular waveguide (E _X component), the position of (A) is a graph showing the electric field distribution, (B) is its Y = 0 [μm] it is a graph showing the E _X component at. It is sectional drawing which shows an example of the dimension of each part of the Y branch waveguide used by the simulation shown to FIG.8 and FIG.9. Is a graph obtained by simulation of an electric field distribution of the branching waveguide with rectangular waveguide (E _X component), (A) is a graph showing the electric field distribution, (B) is its Y = 0 [μm] it is a graph showing the E _X component at the position of. Is a graph obtained by simulation of an electric field distribution of the branching waveguide with half rib waveguide (E _X component), (A) is a graph showing the electric field distribution, (B) is its Y = 0 [μm is a graph showing the E _X component at the position of. It is sectional drawing which shows an example of the dimension of each part of the Y branch waveguide used by the simulation shown in FIG. It is sectional drawing which shows an example of the dimension of each part of the Y branch waveguide used in the simulation shown in FIG. The substrate type | mold optical waveguide element which concerns on the 2nd Embodiment of this invention is shown, (A) is the top view, (B) is sectional drawing in the input waveguide, (C) is sectional drawing in the output waveguide. is there. The substrate type | mold optical waveguide element which concerns on the 3rd Embodiment of this invention is shown, (A) is the top view, (B) is a top view which shows the taper waveguide, (C) is the substrate type | mold optical waveguide element. The top view which shows a modification, (D) is a top view which shows the modification of a taper waveguide. The substrate type | mold optical waveguide element which concerns on the 4th Embodiment of this invention is shown, (A) is the top view, (B) is sectional drawing in the input waveguide, (C) is sectional drawing in the output waveguide. is there. The substrate type | mold optical waveguide element which concerns on the 5th Embodiment of this invention is shown, (A) is the top view, (B) is sectional drawing in the input waveguide, (C) is sectional drawing in the output waveguide. is there. The Mach-Zehnder type | mold optical modulator and its monitor structure which concern on the 6th Embodiment of this invention are shown, (A) is a schematic diagram which shows the example, (B) is a schematic diagram which shows another example. The Mach-Zehnder type | mold optical modulator and its monitor structure which concern on the 7th Embodiment of this invention are shown, (A) is a schematic diagram which shows the example, (B) is a schematic diagram which shows another example. The Mach-Zehnder type output polarization conversion element concerning the 8th Embodiment of this invention is shown, (A) is a schematic diagram which shows the example, (B) is a schematic diagram which shows another example. The dimension of each part in the Y branch waveguide of Example 1 is shown, (A) is a plan view thereof, (B) is a sectional view in the input waveguide, and (C) is a sectional view in the output waveguide. 6 is a graph showing a change in excess loss with respect to the width X of the output waveguide of Example 1 obtained by simulation. It is a graph showing the excess loss at the time of conversion to the TE ₀ mode from the even mode of the first embodiment. Is a graph showing the excess loss at the time of conversion to TE ₁ mode from the odd mode of the first embodiment. 3 is a graph obtained by simulating the electric field distribution in the Y-branch waveguide of Example 1. It is the graph which calculated | required the electric field distribution in case the phase difference in Example 1 is 0 rad by simulation. It is the graph which calculated | required the electric field distribution in case the phase difference in Example 1 is (pi) rad by simulation. The dimension of each part in the Y branch waveguide of the comparative example 1 is shown, (A) is a plan view thereof, (B) is a sectional view of the input waveguide, and (C) is a sectional view of the output waveguide. 6 is a graph showing a change in excess loss with respect to the width X of the output waveguide of Comparative Example 1 obtained by simulation. From even mode of Comparative Example 1 is a graph showing the excess loss at the time of conversion to the TE ₀ mode. From the odd mode of Comparative Example 1 is a graph showing the excess loss at the time of conversion to TE ₁ mode. The dimension of each part in the Y branch waveguide of Example 2 is shown, (A) is a plan view thereof, (B) is a sectional view of the input waveguide, and (C) is a sectional view of the output waveguide. It is the graph which calculated | required the change of the excess loss with respect to the width | variety X of the output waveguide of Example 2 by simulation. A conventional general Y-branch waveguide is shown, (A) is an ideal plan view thereof, (B) is a sectional view thereof, and (C) is a plan view showing an enlarged main portion thereof. The Y branch waveguide of a nonpatent literature 1 is shown, (A) is the top view, (B) is sectional drawing in the input waveguide, (C) is sectional drawing in the output waveguide.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
In all of the following drawings, in order to make each component easy to see, the scale of the size may be changed depending on the component. In addition, the materials, dimensions, and the like exemplified in the following description are merely examples, and the present invention is not necessarily limited thereto, and can be appropriately modified and implemented without departing from the scope of the invention. .

(First embodiment)
<Substrate type optical waveguide device>
First, as a first embodiment of the present invention, for example, a substrate type optical waveguide device 1 shown in FIGS. 1A and 1B will be described. 1A is a plan view showing the substrate-type optical waveguide element 1, and FIG. 1B is a substrate-type optical waveguide element taken along the line segment Z ₁ -Z ₁ shown in FIG. 1A. FIG. FIG. 1C is a cross-sectional view of the substrate-type optical waveguide device 1 taken along line Z ₂ -Z ₂ shown in FIG.

  A substrate type optical waveguide device 1 shown in FIGS. 1A, 1 </ b> B, and 1 </ b> C covers a core 4 that forms a Y-branch waveguide 3 on a substrate 2, covers the core 4, and is more than the core 4. And a clad 5 having a small refractive index. The Y-branch waveguide 3 includes a first waveguide 6 and a second waveguide 7 that are parallel and parallel to each other, and one end of the first waveguide 6 and the second waveguide 7 connected to the other end. 3 waveguides 8.

The substrate type optical waveguide device 1 can be manufactured using, for example, an SOI (Silicon on insulator) wafer made of Si—SiO ₂ —Si. Specifically, the core 4 (Y branch waveguide 3) can be formed by processing the uppermost Si layer of the SOI wafer. When using an SOI wafer, the lower clad 13 becomes Si a substrate 2 in the lowermost layer, it is SiO ₂ in the middle to later.

  The core 4 includes a first rib portion 9 having a rectangular section that forms the first waveguide 6, a second rib portion 10 having a rectangular section that forms the second waveguide 7, and a third conductor. The third rib portion 11 having a rectangular cross section forming the waveguide 8 and the first rib portion 9, the second rib portion 10, and the third rib portion 11 are smaller in thickness (lower height) than the first rib portion 11. And a slab portion 12 shared between the second rib portion 10 and the third rib portion 11.

  The core 4 (the first rib portion 9, the second rib portion 10, the third rib portion 11, and the slab portion 12) is made of a material having a refractive index higher than that of the cladding 5, preferably Si. In the substrate type optical waveguide device 1, the first rib portion 9, the second rib portion 10, the third rib portion 11 and the slab are etched by etching the Si layer on the uppermost layer of the SOI wafer in two stages. The part 12 can be formed integrally.

In the first waveguide 6 and the second waveguide 7, the widths W ₁ and W _{2 of} the first rib portion 9 and the second rib portion 10 are constant in the respective length directions, and the first Linear waveguide in which the distance (width of the slab portion 12) D between the rib portion 9 and the second rib portion 10 is constant in the length direction of the first rib portion 9 and the second rib portion 10. Is forming. That is, in the Y-branch waveguide 3, the first waveguide 6 and the second waveguide 7 are linear first rib portions 9 having the same material and cross-sectional shape (width and thickness (height)). And the second rib portion 10 are formed in parallel with each other. Further, the width _{W 1} of the first rib portion 9 and the width _{W 2} of the second rib portion 10 has the same width _(W 1 = _{W 2).}

The third waveguide 8 has a center O ′ in the width direction between the third rib portion 11 and a center O ′ in the width direction between the first rib portion 9 and the second rib portion 10, and the width _{W 1} of the first rib portion 9, the width _{W 2} of the second rib portion 10, the distance sum of D _(W 1 + W between the first rib portion 9 and the second rib portion 10 _A linear waveguide is formed in which the width W _{3 of} the third rib portion 11 is larger than ( ₂ + D) (that is, the relationship of “W ₃ > W ₁ + W ₂ + D” is satisfied). That is, in the Y branch waveguide 3, the third waveguide 8 is a linear third rib portion 11 having the same material and thickness (height) as the first rib portion 9 and the second rib portion 10. Is formed by. Further, the third rib portion 11 has a pair of protruding portions 11 a and 11 b that protrude by the same width on both sides in the width direction as compared with the first rib portion 9 and the second rib portion 10.

  The slab part 12 is formed between the mutually opposing side surfaces of the first rib part 9 and the second rib part 10. Thus, the first waveguide 6 and the second waveguide 7 are half-rib guides in which the slab portion 12 is provided only on one side in the width direction between the first rib portion 9 and the second rib portion 10. A waveguide is formed.

  The clad 5 has a lower clad 13 and an upper clad 14. The first waveguide 6, the second waveguide 7, and the third waveguide 8 (the first rib portion 9, the second rib portion 10, the third rib portion 11, and the slab portion 12) are formed by a lower cladding. 13 are formed on the surface. The upper clad 14 covers the surface of the lower clad 13 on which the first rib portion 9, the second rib portion 10, the third rib portion 11 and the slab portion 12 are formed.

The clad 5 is made of a material having a refractive index lower than that of the core 4, specifically, SiO ₂ or SiN, an air layer (however, only the upper clad 14 can be applied to the air layer). In the substrate type optical waveguide device 1, the SiO ₂ layer of the SOI wafer described above forms the lower clad 13, and the air layer above it forms the upper clad 14. Further, the upper clad 14 may be formed of an SiO ₂ layer that covers the surface of the lower clad 13.

  The substrate-type optical waveguide element 1 includes a first bending waveguide 15 having one end connected to the other end of the first waveguide 3 and a first bending waveguide 15 having one end connected to the other end of the second waveguide 7. 2 bending waveguides 16.

  The core 4 includes a first rib portion 9A that forms the first bending waveguide 15, a second rib portion 10A that forms the second bending waveguide 16, and the first rib portion 9A and the second rib portion 9A. It has a slab portion 12A shared between the first rib portion 9A and the second rib portion 10A with a thickness (lower height) smaller than that of the rib portion 10A.

  The first bending waveguide 15 and the second bending waveguide 16 are bent in-plane by the first rib portion 9A and the second rib portion 10A having a constant width and thickness (height) in the length direction. It is formed as follows. The first rib portion 9A and the second rib portion 10A are formed continuously with the first rib portion 9 and the second rib portion 10, respectively. That is, the first rib portion 9A is formed in a rectangular cross section with the same width and thickness (height) as the first rib portion 9. On the other hand, the second rib portion 10A is formed in a rectangular cross section with the same width and thickness (height) as the second rib portion 10.

  The first bending waveguide 15 and the second bending waveguide 16 are such that the distance between the first rib portion 9A and the second rib portion 10A is the first waveguide 6 (first rib portion 9). ) And the second waveguide 7 (second rib portion 10) are bent in an S shape with a predetermined curvature and angle so as to continuously decrease toward the side connected to the second waveguide 7 (second rib portion 10).

  The slab portion 12 </ b> A is formed continuously with the slab portion 12 with the same thickness (height) as the slab portion 12. The slab portion 12A is formed between the side surfaces of the first rib portion 9A and the second rib portion 10A facing each other. As a result, the first bending waveguide 15 and the second bending waveguide 16 are half of which the slab portion 12A is provided only on one side in the width direction of the first rib portion 9A and the second rib portion 10A. A rib waveguide is formed.

  The substrate-type optical waveguide device 1 has a fourth waveguide 17 having the other end connected to one end of the third waveguide 8. The core 4 has a third rib portion 11 </ b> A that forms the fourth waveguide 17. Since the fourth waveguide 17 is continuously connected to the straight waveguide 18 having a width different from that of the third waveguide 8, the fourth waveguide 17 is opposite to the side connected to the third waveguide 8. A tapered waveguide (tapered portion) is formed in which the width of the third rib portion 11A continuously decreases. The third rib portion 11 </ b> A is formed continuously with the third rib portion 11 forming the third waveguide 8 and the third rib portion 11 </ b> B forming the straight waveguide 18. That is, the other end of the third rib portion 11A is formed in a rectangular cross section with the same width and thickness (height) as the third rib portion 11. On the other hand, one end of the third rib portion 11A is formed in a rectangular cross section with the same width and thickness (height) as the third rib portion 11B.

  The straight waveguide 18 is a waveguide for connecting the fourth waveguide 17 and an external waveguide (not shown), and has a constant width and thickness (height) in the length direction. The third rib portion 11B is formed linearly.

  In the substrate type optical waveguide device 1 of the present embodiment, the fourth waveguide 17 is not necessarily an essential configuration, and is not necessary when a waveguide having the same width is connected to the third waveguide 8. Further, the linear waveguide 18 is not an essential configuration, and a waveguide having a shape different from that of the linear waveguide 18 (for example, a bent waveguide) can be connected.

  Further, in the substrate type optical waveguide device 1, the above-described first waveguide 6, the second waveguide 7 and the third waveguide are etched by etching the Si layer, which is the uppermost layer of the SOI wafer, in two stages. Together with the first rib portion 9, the second rib portion 10, the third rib portion 11 and the slab portion 12 that form the waveguide 8, the first bent waveguide 15, the second bent waveguide 16, and the fourth The first rib portion 9A, the second rib portion 10A, the third rib portions 11A and 11B, and the slab portion 12A that form the waveguide 17 and the straight waveguide 18 can be integrally formed.

  In the substrate type optical waveguide device 1 of the present embodiment, the butt coupling portion (branch waveguide structure) of the Y branch waveguide 3 from the first waveguide 6, the second waveguide 7, and the third waveguide 8 described above. 31 is configured. Further, in the substrate type optical waveguide device 1 of the present embodiment, the Y branch from the first waveguide 6 and the first bending waveguide 15 and the first waveguide 6 and the first bending waveguide 15 described above. Two input waveguides 32 and 33 of the waveguide 3 are configured, and one output waveguide 34 of the Y-branch waveguide 3 is formed from the third waveguide 8, the fourth waveguide 17, and the straight waveguide 18 described above. It is configured.

In Y-branch waveguide 3, discontinuously bonded with two input waveguides and light TE ₀ mode inputted to the waveguide 32, 33, TE ₀ mode coupling portions 31 butt and a light output from the output waveguide 34 Let Further, the output waveguide 34 has two input waveguides 32, 33 so that the electric field distribution immediately before being connected to the output waveguide 34 of the input waveguides 32, 33 is efficiently coupled to the electric field distribution of the output waveguide 34. It has a wide shape protruding from both sides in the width direction from 33. That is, the width W ₃ of the third rib 11 has a width W ₁ of the first rib portion 9, the width W ₂ of the second rib portion 10, a first rib portion 9 second rib portion It is larger than the sum (W ₁ + W ₂ + D) with the interval D between 10 (W ₃ > W ₁ + W ₂ + D). In this case, by adjusting the width W3 of the _third rib portion 11 according to the electric field distribution on the input side, it is possible to reduce the loss when multiplexing the light.

  Further, in the substrate-type optical waveguide device 1 of the present embodiment, the first rib portion 9 and the second rib 9 are disposed between the two input waveguides 32 and 33 (the first waveguide 6 and the second waveguide 7). By providing a slab portion 12 that is shared between the first rib portion 9 and the second rib portion 10 with a smaller thickness (lower height) than the rib portion 10, low loss multiplexing is possible. It is said.

  That is, when the slab part 12 is provided between the first rib part 9 and the second rib part 10, light is transmitted from the first rib part 9 (or the second rib part 10) to the slab part 12. It will ooze out greatly. On the other hand, since the slab part 12 exists only between the first rib part 9 and the second rib part 10, what is the slab part 12 of the first rib part 9 (or the second rib part 10)? Light does not ooze from the other side.

As a result, in the Y-branch waveguide 4, light confinement is weakened between the adjacent first rib portion 9 and second rib portion 10, and light penetration between them can be increased. it can. As a result, the TE ₀ mode light input to one input waveguide 32 oozes into the other adjacent input waveguide 33 and has an electric field distribution that approaches the center of these two input waveguides 32 and 33. become. For this reason, it is possible to couple TE ₀ mode light having an electric field distribution concentrated in the center of the output waveguide 34 with high efficiency.

  The substrate-type optical waveguide device 1 of the present embodiment has a configuration in which two separate input waveguides 32 and 33 are gradually approached by the first bending waveguide 15 and the second bending waveguide 16 described above. It has become. In this case, light can be gradually leached into adjacent waveguides, and loss due to a sudden change in electric field distribution can be avoided. This facilitates connection with the separated external waveguide.

  In the present embodiment, the first bending waveguide 15 and the second bending waveguide 16 described above are provided. However, it is not always necessary to use two bending waveguides. It is also possible to arrange the linear waveguides obliquely with respect to the light traveling direction so that these two linear waveguides gradually approach each other. Moreover, it is good also as a structure which combined these.

<Y-branch waveguide>
Next, the principle of light multiplexing by the combined waveguide using the Y branch waveguide will be described.
In the following description, portions equivalent to those of the substrate-type optical waveguide device 1 are denoted by the same reference numerals in the drawings and the like as necessary. One input waveguide 32 is referred to as “input waveguide 1”, and the other input waveguide 33 is referred to as “input waveguide 2”.

When the TE ₀ mode light is input to one of the two waveguides 1 and 2 when the two input waveguides 1 and 2 are sufficiently separated from each other, the TE ₀ mode light is transmitted to the adjacent input waveguides 1 and 2. In the cross section, the two guided modes of the even mode with a symmetric electric field distribution and the odd mode with an antisymmetric electric field distribution (collectively called super modes) are superposed with the same power. Expressed.

Here, FIG graph obtained by simulating the electric field distribution of the even mode (E _X component) by TE ₀ mode in the Y-branch waveguide 3 when the input waveguide 1 and 2 adjacent are sufficiently separated 2 ( a), shown (B), the Fig. 3 graph obtained by simulating the electric field distribution of the odd mode _{(E X} component) (a), shown in (B). Incidentally, FIG. 2 (B) and FIG. 3 (B) is a graph showing the _{E X} component at the position of Y = 0 [μm] in FIG. 2 (A) and FIG. 3 (A). In each graph, X represents a coordinate in the width direction and Y represents a coordinate [μm] in the height direction.

Further, FIG. 4 shows the dimensions of each part of the Y-branch waveguide 3 used in this simulation. In this simulation, a finite element method (FEM) is used, the wavelength is 1550 nm, the core 4 (rib portions 9, 10 and the slab portion 12) is Si (refractive index 3.48), and the cladding 5 (lower portion) The clad 13 and the upper clad 14) are made of SiO ₂ (refractive index 1.44), the thickness (height) of the rib portions 9 and 10 is 220 nm, the thickness (height) of the slab portion 12 is 95 nm, The calculation was performed with the width of 10 being 400 nm and the width of the slab portion 12 being 2000 nm.

FIG. 2 (A), the (B) and FIG. 3 (A), (B), the electric field component in the width direction of the Y-branch waveguide 3 only _{(E X} component) is shown. This is because the TE mode is largest dominant E _X component, its features is because mainly expressed in E _X component.

As can be seen from the graphs of FIGS. 2A and 2B and FIGS. 3A and 3B, the electric field when light in the TE ₀ mode is input to the input waveguide 1 (right side in FIG. 4). The distribution is obtained by superimposing the even mode and the odd mode with a phase difference of 0 [rad]. On the other hand, the electric field distribution when TE ₀ mode light is input to the input waveguide 2 (left side in FIG. 4) is such that the even mode and the odd mode are superimposed with a phase difference of π [rad].

  The two input waveguides 1 and 2 having these super modes gradually approach toward the butt coupling portion 31. At this time, the light oozing out from one input waveguide 1 moves to the other input waveguide 2, whereby the super mode changes to a distribution near the center of the two input waveguides 1 and 2.

Next, FIG graph obtained by simulating the electric field distribution of the even mode (E _X component) by TE ₀ mode in the Y-branch waveguide 3 when the input waveguide 1 and 2 adjacent is approaching 5 (A ), (B) in shown, FIG 6 a graph obtained by the electric field distribution of the odd mode _{(E X} component) simulation (a), shown in (B). Incidentally, and FIG. 5 (B) and 6 (B) is a graph showing the _{E X} component at the position of Y = 0 [μm] in FIGS. 5 (A) and 6 (A). Moreover, the dimension of each part of the Y branch waveguide 3 used by this simulation is shown in FIG. The other conditions in this simulation are the same as in the case shown in FIG.

  From the graphs of FIGS. 5A and 5B and FIGS. 6A and 6B, it can be seen that light is collected between adjacent input waveguides in the even mode.

On the other hand, in a Y-branch waveguide using a conventional general rectangular waveguide (see FIG. 35A), the interval between adjacent input waveguides gradually increases along the light traveling direction. It narrows to. At this time, in the even mode, the light continuously changes to the TE ₀ mode light of the output waveguide. On the other hand, in the odd mode, the light continuously changes to the TE ₁ mode light of the output waveguide. The TE ₁ mode represents a mode having the second highest effective refractive index in the TE mode.

Here, FIG. 8 a graph obtained by simulation of an electric field distribution by the TE ₀ mode of a rectangular waveguide _{(E X} component) (A), simulated (B) to show the electric field distribution due to TE ₁ mode _{(E X} component) 9A and 9B show the graphs obtained from the above. Incidentally, and FIG. 8 (B) and FIG. 9 (B) is a graph showing the _{E X} component at the position of Y = 0 [μm] in FIG. 8 (A) and FIG. 9 (A). The dimensions of the rectangular waveguide used in this simulation are shown in FIG. The other conditions in this simulation are the same as in the case shown in FIG.

From the graphs of FIGS. 8A and 8B and FIGS. 9A and 9B, the TE ₀ mode has a symmetrical electric field distribution in the width direction with a peak at the center, and is separated in the TE ₁ mode. It can be seen that it has an antisymmetric electric field distribution in the width direction with two peaks. Such a characteristic of the electric field distribution also holds when the width of the waveguide is changed. When the width of the waveguide changes, the electric field distribution in the TE ₀ mode and the TE ₁ mode also changes in the width direction. become.

In the conventional general Y-branch waveguide, since the two input waveguides 1 and 2 are continuously changed to one output waveguide, conversion from the even mode to the TE ₀ mode of the output waveguide (or The conversion from the odd mode to the TE ₁ mode of the output waveguide) is also performed continuously. Therefore, by combining the Y branch waveguide with a sufficiently long length, in principle, very low loss multiplexing is possible. Such conversion is generally called conversion by adiabatic change.

  On the other hand, like the Y-branch waveguide shown in FIG. 35A and the Y-branch waveguide 3 shown in FIG. 1, it is connected discontinuously to the output waveguide while keeping the distance between the two input waveguides constant. In the case of a Y-branch waveguide, the conversion is performed according to the following principle.

  That is, when the waveguides having different cross sections are connected, the conversion efficiency T indicating the ratio of the waveguide mode having the input side waveguide to the waveguide mode having the output side waveguide is expressed by the following formula (1). It is expressed as (However, when considering the conversion of the TE mode, the TE mode ignores the contribution of other components because the Ex component is the main component.)

Here, the symbols in the above formula (1) are determined as follows. * Represents a complex conjugate, and integration is performed over the entire cross section of the boundary between the two input waveguides and the output waveguide.
Ex ^{input part} : Ex component of TE mode in input waveguide
Ex ^{output unit} : Ex component of TE mode in output waveguide
K: Other constant

Such a transformation is called a butt joint. From the above equation (1), it can be seen that the greater the electric field distribution overlap, the greater the conversion efficiency. In addition, since the even mode of the input waveguide has a symmetrical electric field distribution in the width direction, it does not couple to the TE ₁ mode of the output waveguide, whereas the odd mode of the input waveguide is opposite in the width direction. It can be seen that it has no electric field distribution and is not coupled to the TE ₀ mode of the output waveguide. Therefore, the conversion efficiency of the TE ₀ mode light output from the output waveguide when the TE ₀ mode light is input to one input waveguide is 50% or less. Of this, a loss of 50% is a principle loss. Similarly, the conversion efficiency of the TE ₁ mode light output from the output waveguide when the TE ₀ mode light is input to one of the input waveguides is the same.

In such a Y-branch waveguide in which two input waveguides are discontinuously connected to the output waveguide, it is better that the electric field distributions coincide with each other in order to increase the conversion efficiency. In particular, since the electric field distribution of the even mode of the input waveguide is converted with high efficiency TE ₀ mode output waveguides closer to the center, the distribution of electric field of the even mode of the input waveguide closer to the center It is necessary to have. For this purpose, it is necessary to increase the coupling coefficient of the two input waveguides.

  To increase the coupling coefficient, narrow the width of the two input waveguides, weaken the light confinement, increase the leaching of light to the outside, or narrow the distance between the two input waveguides. That's fine. However, when considering actual fabrication, there are problems such as reproducibility being lowered if the width of the waveguide is too narrow, or depending on the accuracy of lithography, the waveguide as designed in the mask cannot be fabricated. is there. For this reason, there is a minimum waveguide width that can be fabricated. In addition, when the two input waveguides are too close to each other, the reproducibility is deteriorated due to the influence of adjacent waveguides. Furthermore, there is a minimum waveguide interval that can be produced due to lithography and etching. Therefore, there is a limit to increasing the coupling coefficient of the input waveguide.

As a method for increasing the conversion efficiency from the even mode of the input waveguide to the TE ₀ mode of the output waveguide under the above conditions, the Y-branch waveguide described in Non-Patent Document 1 described above is shown in FIG. ), The output waveguide 303 is wider than the two input waveguides 301 and 302 and protrudes on both sides in the width direction. That is, the width of the output waveguide 303 is larger than the sum of the width of the two input waveguides 301 and 302 and the distance between the two input waveguides 301 and 302.

Thereby, when the coupling coefficient of the input waveguides 301 and 302 is not sufficiently increased and the even mode is not sufficiently centered, the electric field distribution of the TE ₀ mode is increased in the width direction by increasing the width of the output waveguide 303. By extending, it is possible to improve the consistency with the even mode.

  However, in the Y branch waveguide described in Non-Patent Document 1, the two input waveguides 301 and 302 are formed by rectangular waveguides. In addition, a flat-plate type quartz waveguide (diffractive index difference between core and clad of 0.3%) is assumed.

  In such a rectangular waveguide, since the refractive index difference between the core and the clad is small, the light confinement becomes weak and a large coupling coefficient can be obtained. On the other hand, in the case of a silicon fine wire waveguide having a large refractive index difference between the core and the clad, light confinement in the core is large. Therefore, further improvement is required to increase the conversion efficiency. (However, the applicable range of the present invention is not limited to a waveguide having a large refractive index difference between the core and the clad, such as a silicon fine wire waveguide, but extends to general optical waveguides by general total reflection.)

(Effects of the present invention)
Next, the effects of the present invention will be described based on the above contents.
In the present invention, a slab portion made of the same material as that of the rib portion and having a smaller thickness (lower height) than the rib portion is provided between two adjacent input waveguides, so that the two input waveguides are semi-ribbed. It is formed by a waveguide. In this case, it is possible to increase the coupling coefficient without increasing the penetration of light into the adjacent input waveguide and changing the width and interval of the input waveguide. Thereby, the conversion efficiency from the even mode of the input waveguide to the TE ₀ mode of the output waveguide can be increased.

Here, the width and spacing of the two input waveguides when constant, the graph obtained by simulating the electric field distribution of the even mode (E _X component) by TE ₀ mode in the branch waveguide using a rectangular waveguide Figure 11 (a), (B) to show a graph obtained by simulating the electric field distribution of the even mode _{(E X} component) by TE ₀ mode in the branch waveguide using a semi rib waveguide FIG. 12 (a), the Shown in (B). Incidentally, FIG. 11 (B) and FIG. 12 (B) is a graph showing the _{E X} component at the position of Y = 0 [μm] in FIG. 11 (A) and FIG. 12 (A). In this simulation, the dimensions of each part of the Y-branch waveguide using the rectangular waveguide are shown in FIG. 13, and the dimensions of each part of the Y-branch waveguide using the half-rib waveguide are shown in FIG. The other conditions in this simulation are the same as in the case shown in FIG.

From the graphs of FIGS. 11A and 11B and FIGS. 12A and 12B, when compared with the same input waveguide width and spacing, the half-rib is more than the Y-branch waveguide using a rectangular waveguide. It can be seen that the Y-branch waveguide using the waveguide has an electric field distribution closer to the center. Therefore, in the Y-branch waveguide using the half-rib waveguide, it is possible to increase the conversion efficiency from the even mode of the input waveguide to the TE ₀ mode of the output waveguide.

Moreover, the following effects 1-7 are mentioned as another effect by this invention.
[Effect 1]
As the effect 1 of the present invention, the wavelength dependency is small. In general, when the wavelength changes, the degree of light confinement changes and the coupling coefficient changes. For example, as the wavelength increases, light leaching from the core increases and the coupling coefficient increases. On the contrary, when the wavelength is shortened, the oozing of light from the core is reduced and the coupling coefficient is reduced. For this reason, conversion efficiency changes with wavelengths. On the other hand, when assuming wavelength division multiplex (WDM) communication, it is preferable that the functional element used therein has a smaller dependency of the conversion efficiency on the wavelength.

  In the Y branch waveguide using the half-rib waveguide of the present invention, the coupling coefficient can be increased as compared with the Y branch waveguide using the conventional rectangular waveguide. Therefore, in the present invention, since the amount of change in the coupling coefficient with respect to the wavelength is small, an effect that the wavelength dependency is small can be obtained.

[Effect 2]
Advantage 2 of the present invention is that it is resistant to manufacturing errors. When the Y-branch waveguide is actually manufactured, the width of the rib part, the height of the rib part changes, or the cross-sectional shape of the rib part has a slight inclination with respect to the value at the design stage. It becomes a shape. In such a case, the degree of light confinement changes as the waveguide shape changes, and the coupling coefficient changes. It is not preferable that the conversion efficiency decreases with respect to such a change.

  On the other hand, in the Y branch waveguide using the half rib waveguide of the present invention, the coupling coefficient can be increased as compared with the Y branch waveguide using the conventional rectangular waveguide. Therefore, according to the present invention, even if the coupling coefficient changes due to a manufacturing error, since a reduction in conversion efficiency is suppressed as compared with the conventional case, an effect that the manufacturing error is strong can be obtained.

[Effect 3]
The effect 3 according to the present invention includes high conversion efficiency from the odd mode to the TE ₁ mode. Since the Y-branch waveguide using the half-rib waveguide of the present invention has a higher coupling coefficient than the Y-branch waveguide using the conventional rectangular waveguide, the even-mode of the input waveguide can be changed to the output waveguide. In addition to the conversion efficiency of the TE waveguide into the TE ₀ mode, the conversion efficiency from the odd mode of the input waveguide to the TE ₁ mode of the output waveguide can be increased. In this case, low loss multiplexing (conversion) is possible by using the device or system utilizing the generation of the TE ₁ mode described later.

[Effect 4]
The effect 4 according to the present invention is that other rib waveguide fabrication processes can be used. The Y-branch waveguide using the semi-rib waveguide of the present invention can be manufactured by etching the Si layer on the uppermost layer of the SOI wafer in two stages, so that the Y-branch using the conventional rectangular waveguide is used. Compared to the waveguide, the manufacturing process is increased. However, since there is no restriction on the height of the slab part, if there are, for example, waveguides provided with slab parts on both sides in the width direction of the rib part (so-called rib waveguides), etc. Since it can be manufactured in a lump, it does not substantially increase the manufacturing process.

[Effect 5]
The effect 5 of the present invention includes a short device length. As in the present invention, in the case of a Y-branch waveguide that is discontinuously connected to the output waveguide while keeping the distance between the two input waveguides constant, Since it is not necessary to ensure a sufficient length to continuously change from the input waveguide to one output waveguide, the device length can be shortened.

In particular, in the conventional general Y-branch waveguide, the shape of the electric field distribution is greatly changed from the even mode to the TE ₀ mode in the portion where the two input waveguides are connected to the output waveguide. Needs to change gradually. On the other hand, the present invention has a great advantage with respect to the device length in that there are few such portions.

[Effect 6]
As the effect 6 of the present invention, it is possible to combine with low loss. When the Y-branch waveguide is actually manufactured, there is a problem that light is scattered due to the influence of the roughness of the side wall of the waveguide, resulting in a loss. This is particularly noticeable in a silicon fine wire waveguide having a small device size. On the other hand, the Y branch waveguide using the half-rib waveguide of the present invention has fewer side walls of the waveguide than the Y branch waveguide using the conventional rectangular waveguide. For this reason, the loss due to the side wall roughness is smaller than in the prior art, and low loss multiplexing is possible.

[Effect 7]
The effect 7 according to the present invention include that which can be combined according to TM ₀ mode. In this case, by calculating the conversion efficiency from the even mode or odd mode by the TM ₀ mode of the two input waveguides to the TM ₀ mode or TM ₁ mode of the output waveguide, the total efficiency by the TM ₀ mode having the desired loss can be calculated. Waves are possible.

(Second Embodiment)
<Substrate type optical waveguide device>
Next, a substrate type optical waveguide device 101A shown in FIGS. 15A, 15B, and 15C will be described as a second embodiment of the present invention. 15A is a plan view showing the substrate type optical waveguide device 101A, and FIG. 15B is a substrate type optical waveguide device according to the line segment Z ₁ -Z ₁ shown in FIG. 15A. It is sectional drawing of 101A. FIG. 15C is a cross-sectional view of the substrate-type optical waveguide element 101A along the line segment Z ₂ -Z ₂ shown in FIG. Moreover, in the following description, about the site | part equivalent to the board | substrate type | mold optical waveguide element 1 shown to the said FIG. 1 (A), (B), (C), while omitting description, the same code | symbol shall be attached | subjected in drawing. To do.

  In the substrate-type optical waveguide device 101A shown in FIGS. 15A, 15B, and 15C, the side surfaces of the first rib portions 9 and 9A and the second rib portions 10 and 10A opposite to each other are opposite to each other. The slab portion 12B is continuously provided on the side surface. The rest of the configuration is basically the same as that of the substrate-type optical waveguide device 1. That is, the two input waveguides 32 and 33 of the Y-branch waveguide 3 have slab portions 12, 12 A, and 12 B on both sides in the width direction of the first rib portions 9, 9 A and the second rib portions 10, 10 A. By being provided, a rib waveguide is formed.

  In the substrate type optical waveguide device 101A of the present embodiment, the same effects as those of the substrate type optical waveguide device 1 can be obtained. Further, when the input waveguides 32 and 33 are formed of rib waveguides, the coupling coefficient between the two input waveguides 32 and 33 is higher than that of the rectangular waveguide, and therefore the loss is reduced. On the other hand, since the rib waveguide has slab portions on both sides in the width direction of the rib portion, light oozes not only into the slab portion between the rib portions but also into the slab portion on the opposite side. . Therefore, in the case of the rib waveguide, the coupling coefficient between the two input waveguides 32 and 33 is lower than that in the case of the half-rib waveguide, so that the loss is increased. However, in the case of a rib waveguide, since the waveguide has few side walls, the influence of loss due to the side wall roughness is small. This is advantageous over the semi-rib waveguide.

(Third embodiment)
<Substrate type optical waveguide device>
Next, a substrate type optical waveguide device 101B shown in FIGS. 16A and 16B will be described as a third embodiment of the present invention. 16A is a plan view showing the substrate-type optical waveguide device 101B, and FIG. 16B is a plan view showing the tapered waveguide 102 provided in the substrate-type optical waveguide device 101B. Moreover, in the following description, about the site | part equivalent to the board | substrate type | mold optical waveguide element 1 shown to the said FIG. 1 (A), (B), (C), while omitting description, the same code | symbol shall be attached | subjected in drawing. To do.

  The substrate-type optical waveguide device 101B of this embodiment has a configuration in which a tapered waveguide 102 is connected to a first bent waveguide 15 and a second bent waveguide 16, respectively. The rest of the configuration is basically the same as that of the substrate-type optical waveguide device 1.

  The tapered waveguide 102 is converted from a rib waveguide to a half-rib waveguide (or from a half-rib waveguide to a rib guide) by continuously changing the waveguide structure between the half-rib waveguide and the rectangular waveguide. A waveguide to be converted) is formed.

  Specifically, the tapered waveguide 102 includes a first slab portion 12C continuously provided on a side surface of the first rib portion 9A facing the second rib portion 10A, and a second rib portion. The first rib portion 9 </ b> A of 10 </ b> A and the second slab portion 12 </ b> D provided continuously on the side surface facing the 10 </ b> A. The first slab part 12C and the second slab part 12D are provided continuously to the slab part 12A, and the respective widths are continuously increased toward the slab part 12A.

  The first slab part 12C and the second slab part 12D have a constant taper angle and gradually increase the width from the middle part of the first rib part 9A and the second rib part 10A. The slab part 12A is extended.

  In the substrate type optical waveguide device 101B of the present embodiment, the same effects as those of the substrate type optical waveguide device 1 can be obtained. Further, in the substrate type optical waveguide device 101B of the present embodiment, by using the tapered waveguide 102, the widths of the slab portions 12C and 12D can be continuously changed between the rectangular waveguide and the half-rib waveguide. it can. This facilitates the connection between the rectangular waveguide and the half-rib waveguide.

  A modification of the substrate type optical waveguide device 101B is shown in FIG. In the substrate-type optical waveguide device 101B shown in FIG. 16C, the first bending waveguide 15 and the second bending waveguide 16 are formed by bending at a right angle in a direction away from each other with a predetermined curvature. . Also in this case, the first slab part 12C and the second slab part 12D are provided continuously to the slab part 12A, and the respective widths are continuously increased toward the slab part 12A.

  On the other hand, when a tapered waveguide 103 as shown in FIG. 16D is used, the waveguide structure is converted from a rib waveguide to a half-rib waveguide, or converted from a half-rib waveguide to a rib waveguide. A waveguide structure is obtained.

(Fourth embodiment)
<Substrate type optical waveguide device>
Next, as a fourth embodiment, a substrate type optical waveguide device 101C shown in FIGS. 17A, 17B, and 17C will be described. FIG. 17A is a plan view showing the substrate type optical waveguide device 101C, and FIG. 17B is a substrate type optical waveguide device according to the line segment Z ₁ -Z ₁ shown in FIG. It is sectional drawing of 101C. FIG. 17C is a cross-sectional view of the substrate-type optical waveguide element 101C along the line segment Z ₂ -Z ₂ shown in FIG. Moreover, in the following description, about the site | part equivalent to the board | substrate type | mold optical waveguide element 1 shown to the said FIG. 1 (A), (B), (C), while omitting description, the same code | symbol shall be attached | subjected in drawing. To do.

  In the substrate-type optical waveguide device 101C shown in FIGS. 17A, 17B, and 17C, the slab portions 12E are continuously provided on the side surfaces on both sides of the third rib portion 11. The rest of the configuration is basically the same as that of the substrate-type optical waveguide device 1. That is, one output waveguide 34 of the Y branch waveguide 3 forms a rib waveguide by providing the slab portions 12E on the side surfaces on both sides of the third rib portion 11.

  In the substrate type optical waveguide device 101C of the present embodiment, the same effects as those of the substrate type optical waveguide device 1 can be obtained. Further, when the output waveguide 34 is formed of a rib waveguide, the output waveguide 34 can be easily connected to an external rib waveguide, and conversion of the waveguide structure becomes unnecessary, so that the device length can be shortened. . Moreover, the loss due to the above-described side wall roughness can be reduced.

(Fifth embodiment)
<Substrate type optical waveguide device>
Next, as a fifth embodiment, a substrate type optical waveguide device 101D shown in FIGS. 18A, 18B, and 18C will be described. 18A is a plan view showing the substrate type optical waveguide device 101D, and FIG. 18B is a substrate type optical waveguide device according to the line segment Z ₁ -Z ₁ shown in FIG. 18A. It is sectional drawing of 101D. FIG. 18C is a cross-sectional view of the substrate-type optical waveguide element 101D along the line segment Z ₂ -Z ₂ shown in FIG. Moreover, in the following description, about the site | part equivalent to the board | substrate type | mold optical waveguide element 1 shown to the said FIG. 1 (A), (B), (C), while omitting description, the same code | symbol shall be attached | subjected in drawing. To do.

  In the substrate-type optical waveguide device 101D shown in FIGS. 18A, 18B, and 18C, the side surfaces of the first rib portions 9 and 9A and the second rib portions 10 and 10A that are opposite to each other are opposite to each other. The slab portion 12B and the slab portions 12E are provided continuously on the side surfaces on both sides of the third rib portion 11, respectively. The rest of the configuration is basically the same as that of the substrate-type optical waveguide device 1. That is, the two input waveguides 32 and 33 of the Y-branch waveguide 3 have slab portions 12, 12 A, and 12 B on both sides in the width direction of the first rib portions 9, 9 A and the second rib portions 10, 10 A. By being provided, a rib waveguide is formed. Further, one output waveguide 34 of the Y branch waveguide 3 forms a rib waveguide by providing slab portions 12E on both side surfaces of the third rib portion 11.

  In the substrate type optical waveguide device 101D of the present embodiment, the same effects as those of the substrate type optical waveguide device 1 can be obtained. Further, when the input waveguides 32 and 33 and the output waveguide 34 are formed of rib waveguides, the same effects as those of the substrate type optical waveguide device 101A and the substrate type optical waveguide device 101C can be obtained.

(Sixth embodiment)
Next, as a sixth embodiment, a Mach-Zehnder optical modulator (MZM) 50A shown in FIG. 19A and its monitor structure, and a Mach-Zehnder optical modulator 50B shown in FIG. The monitor structure will be described. FIG. 19A is a schematic diagram showing a Mach-Zehnder optical modulator 50A and its monitor structure. FIG. 19B is a schematic diagram showing a Mach-Zehnder optical modulator 50B and its monitor structure.

  A Mach-Zehnder optical modulator 50A shown in FIG. 19A includes an input unit 51 to which light is input, a demultiplexing unit 52 that demultiplexes light input from the input unit 51, and demultiplexing by the demultiplexing unit 52. Two phase modulation units 53 and 54 that phase-modulate the modulated light, a multiplexing unit 55 that combines the light phase-modulated by the two phase modulation units 53 and 54, and a multiplexing unit 55 And an output unit 56 that outputs light.

  In the Mach-Zehnder type optical modulator 50A shown in FIG. 19A, by using the substrate type optical waveguide device 1 including the Y branch waveguide 3 for the multiplexing unit 55, low loss multiplexing is possible. ing.

Further, in the Mach-Zehnder optical modulator 50A shown in FIG. 19A, the high-order mode splitter 57 that separates the TE _1- mode light and is separated by the high-order mode splitter is positioned after the output unit 56. A monitor structure in which a photodetector (PD: Photo Detector) 58 for detecting light in TE ₁ mode is arranged is employed.

In this monitor structure, TE ₀ mode light (represented by an arrow TE ₀ in FIG. 19A) and TE ₁ mode light (represented by an arrow TE ₁ in FIG. 19A) are: When simultaneously output from the output unit 56, the light in the TE ₁ mode is separated and detected. Specifically, in this monitoring structure, by utilizing the fact that the TE ₁ mode of light occurs when the light is canceled by the optical modulator 50A, after separation of the light of the TE ₁ mode in higher mode splitter 57 The light detector 58 detects the separated TE _1- mode light and converts it into an electrical signal, so that the driving condition of the optical modulator 50A can be monitored.

In the Mach-Zehnder optical modulator 50A shown in FIG. 19A, by using the substrate-type optical waveguide device 1 including the Y branch waveguide 3 for the multiplexing unit 55, not only the TE ₀ mode light, The TE ₁ mode light can also be generated with low loss. Therefore, in this monitor structure, since the power of the TE ₁ mode light detected by the photodetector 58 is increased, more accurate monitor control resistant to noise can be efficiently performed.

  A Mach-Zehnder optical modulator 50B shown in FIG. 19B includes a first input unit 51A to which light is input and a first demultiplexing unit 52A that demultiplexes the light input from the first input unit 51A. And two phase modulation units 53 and 54 that phase-modulate the light demultiplexed by the first demultiplexing unit 52A, and the first light that is phase-modulated by the two phase modulation units 53 and 54. A first light modulation unit 59 and a second light modulation unit 60 each including a multiplexing unit 55A and a first output unit 56A that outputs the light combined by the first multiplexing unit 55A are provided. ing.

  Also, the Mach-Zehnder type optical modulator 50B shown in FIG. 19B is positioned in front of the first light modulation unit 59 and the second light modulation unit 60, and the second input unit 51B to which light is input. And demultiplexing the light input from the second input unit 51B to the first input unit 51A side of the first light modulation unit 59 and the first input unit 51A side of the second light modulation unit 60. Light output from the first output unit 56 </ b> A of the first light modulation unit 59, located after the second demultiplexing unit 52 </ b> B, the first light modulation unit 59, and the second light modulation unit 60. And the light output from the first output unit 56A of the second light modulation unit 60 is combined with the second combining unit 55B, and the light combined with the second combining unit 55B is output. Second output unit 56B.

  Further, the Mach-Zehnder type optical modulator 50B shown in FIG. 19B is one of the first light modulation unit 59 and the second light modulation unit 60 (the second light modulation unit 60 in the present embodiment). The phase adjusting unit 61 for adjusting the phase difference of the light output from the first multiplexing unit 55A is provided.

  In the Mach-Zehnder optical modulator 50B shown in FIG. 19B, either the first multiplexing unit 55A or the second multiplexing unit 55B (the second multiplexing unit 55B in the present embodiment) is connected to the above. By using the substrate type optical waveguide device 1 provided with the Y branching waveguide 3, low loss multiplexing is possible.

  Further, in the Mach-Zehnder type optical modulator 50B shown in FIG. 19B, similarly to the Mach-Zehnder type optical modulator 50A shown in FIG. 19A, the Mach-Zehnder type optical modulator 50B is located at the subsequent stage of the second output unit 56B. A monitor structure in which a mode splitter 57 and a photodetector 58 are arranged is adopted.

In this monitor structure, TE ₀ mode light (represented by an arrow TE ₀ in FIG. 19B) and TE ₁ mode light (represented by an arrow TE ₁ in FIG. 19B). When simultaneously output from the second output unit 56B, the TE ₁ mode light is separated and detected. Specifically, in this monitoring structure, by utilizing the fact that the TE ₁ mode of light occurs when the light is canceled by the optical modulator 50B, after separation of the light of the TE ₁ mode in higher mode splitter 57 The light detector 58 detects the separated TE _1- mode light and converts it into an electric signal, so that the driving condition of the optical modulator 50B can be monitored.

In the Mach-Zehnder optical modulator 50B shown in FIG. 19B, either the first multiplexing unit 55A or the second multiplexing unit 55B (the second multiplexing unit 55B in the present embodiment) is connected to the above. By using the substrate-type optical waveguide device 1 including the Y branch waveguide 3, not only TE ₀ mode light but also TE ₁ mode light can be generated with low loss. Therefore, in this monitor structure, since the power of the TE ₁ mode light detected by the photodetector 58 is increased, more accurate monitor control resistant to noise can be efficiently performed.

(Seventh embodiment)
Next, as a seventh embodiment, a Mach-Zehnder optical modulator 70A shown in FIG. 20A and its monitor structure, and a Mach-Zehnder optical modulator 70B shown in FIG. 20B and its monitor structure will be described. . FIG. 20A is a schematic diagram showing a Mach-Zehnder optical modulator 70A and its monitor structure. FIG. 20B is a schematic diagram showing a Mach-Zehnder optical modulator 70B and its monitor structure. In the following description, parts equivalent to those of the Mach-Zehnder type optical modulators 50A and 50B shown in FIGS. 20A and 20B are not described and are given the same reference numerals in the drawings.

  A Mach-Zehnder optical modulator 70A shown in FIG. 20A further includes a phase adjustment unit 61 that adjusts the phase difference of the light output from the multiplexing unit 55 in addition to the configuration of the optical modulator 50A. Yes.

  In the Mach-Zehnder optical modulator 70A shown in FIG. 20A, by using the substrate type optical waveguide device 1 including the Y branch waveguide 3 for the multiplexing unit 55, low loss multiplexing is possible. ing.

Further, in the Mach-Zehnder type optical modulator 70A shown in FIG. 20A, a high-order polarization conversion unit 71 that is positioned after the output unit 56 and converts TE ₁ mode light into TM ₀ mode light; polarization beam splitter for separating the light of the TE ₀ mode: and (PBS Polarizing beam splitter) 72, a monitor structure in which a photodetector 58 disposed to detect the light of the polarization beam splitter 72 in an isolated TE ₀ mode Adopted.

In this monitor structure, TE ₀ mode light (represented by an arrow TE ₀ in FIG. 20A) and TE ₁ mode light (represented by an arrow TE ₁ in FIG. 20A) are: When simultaneously output from the output unit 56, the TE ₁ mode light is converted into TM ₀ mode light (represented by the arrow TM ₀ in FIG. 20A), and then the TE ₀ mode light is separated. To detect.

Specifically, in this monitoring structure, by utilizing the fact that the TE ₁ mode of light occurs when the light is canceled by the optical modulator 70A, the light of the TE ₁ mode high-order polarization conversion unit 71 TM It converts to ₀ mode light and outputs. As the high-order polarization converter 71, for example, the following reference [1] (Daoxin Dai, et al., “Novel concept for ultracompact polarization splitter-rotator based on silicon nanowires,” Optics Express, Vol. 11, 2011, 10940-10949) and a high-order polarization conversion element disclosed in Japanese Patent Application No. 2013-135490 can be used.

In the high-order polarization conversion unit 71, when light in the TE ₀ mode is input, conversion into another waveguide mode is not performed. Therefore, the TE ₀ mode light generated by the optical modulator 70A is output as it is to the subsequent stage of the high-order polarization converter 71. Thereafter, the TE ₀ mode light and the TM ₀ mode light output from the high-order polarization converter 71 are separated by the polarization beam splitter 72. The separated light of TE ₀ mode is detected by the photodetector 58 and converted into an electric signal, so that the driving condition of the optical modulator 70A can be monitored.

In the Mach-Zehnder optical modulator 70A shown in FIG. 20A, by using the substrate-type optical waveguide device 1 including the Y-branch waveguide 3 for the multiplexing unit 55, not only TE ₀ mode light, The TE ₁ mode light can also be generated with low loss. Therefore, in this monitor structure, the power of the monitoring TE ₀ mode light output from the high-order polarization converter 71 and the signal TM ₀ mode light can be maintained high.

  A Mach-Zehnder optical modulator 70B shown in FIG. 20B has the configuration of the optical modulator 50B. In the Mach-Zehnder optical modulator 70A shown in FIG. 20A, either the first multiplexing unit 55A or the second multiplexing unit 55B (the second multiplexing unit 55B in the present embodiment) is connected to the above By using the substrate type optical waveguide device 1 provided with the Y branching waveguide 3, low loss multiplexing is possible.

  Further, in the Mach-Zehnder type optical modulator 70B shown in FIG. 20A, similarly to the Mach-Zehnder type optical modulator 70A shown in FIG. A monitor structure in which a polarization converter 71, a polarization beam splitter 72, and a photodetector 58 are arranged is employed.

In this monitor structure, TE ₀ mode light (represented by an arrow TE ₀ in FIG. 20B) and TE ₁ mode light (represented by an arrow TE ₁ in FIG. 20B) are: When simultaneously output from the output unit 56, the TE ₁ mode light is converted into TM ₀ mode light (represented by the arrow TM ₀ in FIG. 20B), and then the TE ₀ mode light is separated. To detect.

Specifically, in this monitoring structure, by utilizing the fact that the TE ₁ mode of light occurs when the light is canceled by the optical modulator 70B, the light of the TE ₁ mode high-order polarization conversion unit 71 TM It converts to ₀ mode light and outputs. In the high-order polarization conversion unit 71, when light in the TE ₀ mode is input, conversion into another waveguide mode is not performed. Therefore, the TE ₀ mode light generated by the optical modulator 70B is output as it is to the subsequent stage of the high-order polarization converter 71. Thereafter, the TE ₀ mode light and the TM ₀ mode light output from the high-order polarization converter 71 are separated by the polarization beam splitter 72. The separated light of the TE ₀ mode is detected by the photodetector 58 and converted into an electric signal, so that the driving condition of the optical modulator 70B can be monitored.

In the Mach-Zehnder optical modulator 70B shown in FIG. 20 (B), either the first multiplexing unit 55A or the second multiplexing unit 55B (the second multiplexing unit 55B in the present embodiment) is connected to the above. By using the substrate-type optical waveguide device 1 including the Y branch waveguide 3, not only TE ₀ mode light but also TE ₁ mode light can be generated with low loss. Therefore, in this monitor structure, the power of the monitoring TE ₀ mode light output from the high-order polarization converter 71 and the signal TM ₀ mode light can be maintained high.

(Eighth embodiment)
Next, as an eighth embodiment, a Mach-Zehnder output polarization conversion element 80A shown in FIG. 21A and a Mach-Zehnder output polarization conversion element 80B shown in FIG. 21B will be described. FIG. 21A is a schematic diagram showing a Mach-Zehnder output polarization conversion element 80A. FIG. 21B is a schematic diagram showing a Mach-Zehnder output polarization conversion element 80B.

  A Mach-Zehnder type output polarization conversion element 80A shown in FIG. 21A includes an input unit 81 to which light is input, a demultiplexing unit 82 for demultiplexing light input from the input unit 81, and a demultiplexing unit 82. A Mach-Zehnder Interferometer (MZI) 85 having a multiplexing unit 83 that combines the demultiplexed light and an output unit 84 that outputs the light combined by the multiplexing unit 83 is provided. .

Further, the Mach-Zehnder type output polarization conversion element 80A shown in FIG. 21A includes TE ₀ mode light (represented by an arrow TE ₀ in FIG. 21A) and TE ₁ mode light (FIG. 21 (A)). A) is represented by the arrow TE ₁ in FIG. 5B), and a high-order polarization conversion unit 86 that converts TE ₁ mode light into TM ₀ mode light, and a demultiplexing unit 82. And a phase adjusting unit 87 that adjusts the phase difference of any one of the lights demultiplexed in.

In the Mach-Zehnder type output polarization conversion element 80A shown in FIG. 21A, TE ₀ mode light (indicated by the arrow TE ₀ in FIG. 21A) is input from the input unit 81 of the Mach-Zehnder interferometer 85. The polarization mode of the light output from the output unit 84 of the Mach-Zehnder interferometer 85 is adjusted by the phase adjustment unit 87, and light in the TE ₀ mode (represented by the arrow TE ₀ in FIG. 21A) and TM _0. mode light can be switched to any of the (. represented by arrow TM ₀ in FIG. 21 (a)).

In the Mach-Zehnder output polarization conversion element 80A shown in FIG. 21A, by using the substrate-type optical waveguide element 1 including the Y branch waveguide 3 for the multiplexing unit 83, TE ₀ mode light and TM It is possible to generate _zero- mode light with low loss. Therefore, in this output polarization conversion element 80A, the power of TE ₀ mode light and TM ₀ mode light that are finally output can be increased.

  A Mach-Zehnder output polarization conversion element 80B shown in FIG. 21B has two phase modulators 88 and 89 for phase-modulating the light demultiplexed by the demultiplexing unit 82 in the configuration of the Mach-Zehnder interferometer 85. By doing so, an optical modulation unit (MZM) 90 is configured.

In the Mach-Zehnder type output polarization conversion element 80B shown in FIG. 21B, TE ₀ mode light (indicated by the arrow TE ₀ in FIG. 21A) is input from the input unit 81 of the light modulation unit 90. The polarization mode of the light output from the output unit 84 of the light modulation unit 90 is adjusted by the phase adjustment unit 87, and light in the TE ₀ mode (represented by the arrow TE ₀ in FIG. 21A) and TM _0. mode light can be switched to any of the (. represented by arrow TM ₀ in FIG. 21 (a)).

In the Mach-Zehnder type output polarization conversion element 80B shown in FIG. 21 (B), by using the substrate type optical waveguide element 1 having the Y branch waveguide 3 for the multiplexing unit 83, TE ₀ mode light and TM It is possible to generate _zero- mode light with low loss. Therefore, in this output polarization conversion element 80B, the power of the TE ₀ mode light and the TM ₀ mode light that are finally output can be increased.

In the Y branch waveguide 3, the loss between the TE ₀ mode light and the TM ₀ mode light can be adjusted by adjusting the width of the output waveguide 34. For this reason, it is possible to eliminate the power imbalance between the TE ₀ mode light and the TM ₀ mode light that are finally output in consideration of the loss in the output polarization conversion elements 80A and 80B. .

(Other embodiments)
In addition, this invention is not necessarily limited to the thing of the said embodiment, A various change can be added in the range which does not deviate from the meaning of this invention.
For example, in the above embodiment, the Y branching waveguide 3 is not limited to being used as a multiplexing waveguide for multiplexing light, but the Y branching waveguide 3 may be used as a demultiplexing waveguide for demultiplexing light. Is possible. That is, in the Y branch waveguide 3, one input waveguide is constituted by the third waveguide 8, the fourth waveguide 17, and the straight waveguide 18, and the first waveguide 6 and the first bending guide are formed. Two output waveguides may be configured from the waveguide 15, the first waveguide 6, and the first bending waveguide 15. The Y branch waveguide 3 can perform low-loss demultiplexing even when used as a demultiplexing waveguide.

  Hereinafter, the effects of the present invention will be made clearer by examples. In addition, this invention is not limited to a following example, In the range which does not change the summary, it can change suitably and can implement.

Example 1
Example 1 is an example corresponding to the substrate-type optical waveguide device 1 of the first embodiment. The dimensions of each part in the Y-branch waveguide 3 are shown in FIGS. 22 (A), (B), and (C). It is shown. FIG. 22A is a plan view showing the dimensions of each part. FIG. 22B is a cross-sectional view of the input waveguides 32 and 33 along the line segment Z ₃ -Z ₃ in FIG. FIG. 22C is a cross-sectional view of the output waveguide 34 taken along the line segment Z ₄ -Z ₄ in FIG.

First, the output waveguide 34 (the third waveguide 8) when the even mode and the odd mode of the two input waveguides 32 and 33 are converted into the TE ₀ mode and the TE ₁ mode of the output waveguide 34, respectively. FIG. 23 shows a graph in which the change in excess loss (= −conversion efficiency [dB]) with respect to the width X [nm] is determined by simulation. The conditions for this simulation are the same as those shown in FIG. 4 except that the dimensions shown in FIGS. 22A, 22B, and 22C are used. In this simulation, the calculation is performed using the finite element method (FEM), and the conversion efficiency by the butt coupling unit 31 is calculated (by the accurate calculation that does not approximate the equation (1)).

In the graph in FIG. 23, TE ₀ indicated by a solid line represents excess loss during conversion from the even mode to the TE ₀ mode, and TE ₁ indicated by a broken line represents that during conversion from the odd mode to the TE ₁ mode. Represents excess loss. Further, when X = 1100 nm, there is no pair of projecting portions 11a and 11b of the output waveguide 34.

From the graph shown in FIG. 23, it can be seen that as the width X of the output waveguide 34 increases from X = 1100 nm, the overlap of the electric field distribution increases, and the excessive loss gradually decreases and takes a certain minimum value. . From this, it can be seen that when the width X of the output waveguide 34 is increased, the TE ₀ mode electric field distribution of the output waveguide 34 is widened, whereby the excess loss is gradually increased. Therefore, in order to reduce excess loss, it is preferable that the width X of the output waveguide 34 is made larger than 1100 nm (that is, a state where there is a pair of projecting portions 11a and 11b) and not too large.

Therefore, consider the case where the excessive loss in the TE ₀ mode takes a minimum value (X = about 1500 nm). The excess loss in this case was 0.24 dB in the conversion from the even mode to the TE ₀ mode, and 0.32 dB in the conversion from the odd mode to the TE ₁ mode.

Note that the TE ₁ mode minimum value does not match the X value when the TE ₀ mode takes the minimum value. This is because, in the TE ₁ mode, the optimum X value is such that the two peaks of the odd mode of the two input waveguides 32 and 33 match the two peaks of the TE ₁ mode of the output waveguide 34. On the other hand, in the TE ₀ mode, the X value is determined in consideration of the spread from the center of the even mode and its degree, and the reason for the loss reduction is different. For this reason, when it is desired to operate not only in the TE ₀ mode but also in the TE ₁ mode, it is necessary to appropriately adjust the X value according to the required specifications.

Next, regarding the wavelength dependence of excess loss when X = 1500 nm, graphs obtained by simulation of changes in excess loss with respect to wavelength are shown in FIGS. FIG. 24 is a graph showing excess loss at the time of conversion from the even mode to the TE ₀ mode, and FIG. 25 is a graph showing excess loss at the time of conversion from the odd mode to the TE ₁ mode.

In the graphs shown in FIGS. 24 and 25, TE ₀ and TE ₁ indicated by solid lines represent cases where the change in the waveguide width is 0 nm (as designed), and TE ₀ and TE ₁ indicated by dotted lines are The waveguide width is changed by −30 nm due to a manufacturing error, and TE ₀ and TE ₁ indicated by broken lines indicate the case where the waveguide width is changed by +30 nm due to a manufacturing error. The change in the waveguide width assumes a case where all the waveguide widths constituting the Y-branch waveguide change symmetrically in the width direction while maintaining the center position in the width direction. This is a typical manufacturing error when manufacturing by lithography and etching.

From the graphs shown in FIG. 24 and FIG. 25, the change in excess loss at the time of conversion from the even mode to the TE ₀ mode was 0.20 dB in the wavelength range of 1480 to 1680 nm. When the waveguide width fluctuated by ± 30 nm, the change in excess loss during conversion from the even mode to the TE ₀ mode was 0.027 dB at maximum in the wavelength range of 1480 to 1680 nm.

  This result is very small compared to the case of the Y branch waveguide using the rectangular waveguide described in Non-Patent Document 1 shown in Comparative Example 1 described later (see FIG. 36A).

  In the first embodiment, the conversion efficiency at the butt coupling portion 31 is calculated, but actually, it is necessary to calculate the loss of the bent waveguide, the tapered portion, and the like. However, as a general argument, it is possible to sufficiently reduce the loss by increasing the curvature of the bent waveguide or increasing the taper length of the taper portion. Therefore, in this embodiment, the calculation is limited to the butt coupling portion 31 that most affects the loss.

Next, the excess loss of the TE ₀ mode light output from the output waveguide 34 when the TE ₀ mode light is input to one input waveguide 32 with X = 1500 nm was calculated by simulation. FIG. 26 shows a graph representing the electric field distribution at that time. The wavelength was 1580 nm, and the FDTD (Finite-difference time-domain) method was used for the calculation.

The excess loss (excluding the 3 dB principle loss) of the TE ₀ mode light output from the output waveguide 34 was 0.23 dB, and the excess loss at the butt coupling portion 31 was 0.21 dB. From this, it can be understood that the excess loss of the butt coupling portion 31 is dominant, and this value indicates the total loss (however, the exact match is not seen due to the calculation error). Therefore, in the graphs shown in FIGS. 23, 24 and 25, the excess loss of the butt coupling portion 31 can be regarded as the total loss.

In the graph shown in FIG. 26, not only the TE ₀ mode light but also the TE ₁ mode light is generated in the butt coupling unit 31, and thus the electric field distribution after the combination is TE ₀ mode light and TE ₁ mode light. It is a superposition of mode light and presents a complex aspect.

Next, in order to show the electric field distribution after the combination more clearly, the electric field distribution when X = 1500 nm and the phase difference when the TE ₀ mode light is input to the two input waveguides 32 and 33 is 0 rad is shown. FIG. 27 shows a graph obtained by simulation, and FIG. 28 shows a graph obtained by simulation of the electric field distribution when the phase difference is π rad.

From the graph shown in FIG. 27, when the phase difference is 0 rad, the phases of the odd modes input to the two input waveguides 32 and 33 are opposite to each other, and thus cancel each other. On the other hand, since the phases of the even mode are in phase, they strengthen each other. As a result, in the output waveguide 34, TE ₁ mode light is not output, and the combined TE ₀ mode light can be clearly seen. On the other hand, from the graph shown in FIG. 28, when the phase difference is π rad, TE ₀ mode light is not output in the output waveguide 34, and the combined TE ₁ mode light can be clearly seen.
As described above, it has been clarified from the result of Example 1 that low-loss multiplexing is possible.

(Comparative Example 1)
Comparative Example 1 is a comparative example corresponding to a Y-branch waveguide using a rectangular waveguide described in Non-Patent Document 1, and the dimensions of each part in this Y-branch waveguide are shown in FIGS. ), As shown in (C). FIG. 29A is a plan view showing the dimensions of each part. FIG. 29B is a cross-sectional view of the input waveguide taken along line Z ₃ ′ -Z ₃ ′ in FIG. 29A showing the dimensions of each part. FIG. 29C is a cross-sectional view of the output waveguide taken along line Z ₄ ′ -Z ₄ ′ in FIG. 29A showing the dimensions of the respective parts. That is, the dimension of each part of the Y branch waveguide in Comparative Example 1 is the same as the dimension of each part of the Y branch waveguide in Example 1. The excess loss is calculated using the butt coupling portion 31.

  First, similarly to the case shown in FIG. 23, for the Y-branch waveguide using the rectangular waveguide described in Non-Patent Document 1, a change in excess loss with respect to the width X [nm] of the output waveguide is obtained by simulation. The graph is shown in FIG. Note that the way of viewing the graph shown in FIG. 30 is the same as that shown in FIG.

From the graph shown in FIG. 30, the same tendency as the graph shown in FIG. 23 was observed in the change of excess loss with respect to the width X [nm] of the output waveguide. Specifically, when X = about 1850 nm, the TE ₀ mode excess loss has a minimum value, and the excess loss at this time is 0.63 dB in the conversion from the even mode to the TE ₀ mode, and from the odd mode to the TE _1. Conversion to mode resulted in 0.81 dB.

  From this result, it can be seen that the Y-branch waveguide of Example 1 can be converted with lower loss loss than the Y-branch waveguide of Comparative Example 1. This is because the Y branch waveguide of Example 1 has a larger electric field distribution with the coupling coefficient of the two input waveguides being larger and the even mode being closer to the center.

Further, the X value at which the excess loss in the TE ₀ mode is minimized is about 1500 nm in Example 1, whereas it is as large as about 1850 nm in Comparative Example 1. This is because the coupling in the rectangular waveguide is weak and the electric field distribution of the even mode is distributed at a point spread from the center. Therefore, when the width of the output waveguide is widened and the TE ₀ mode distribution is widened, This is because bonding increases.

On the other hand, in the odd mode, the minimum X value due to the difference in the electric field distribution does not differ as much as the even mode between Example 1 and Comparative Example 1. For this reason, the X value at which the excess loss in the TE ₀ mode and the TE ₁ mode is minimized is larger in Comparative Example 1 than in Example 1. Therefore, in the first embodiment, it is possible to convert the light in the TE ₀ mode and the TE ₁ mode to a lower loss than in the first comparative example (see the above effect 3).

Next, similarly to the case shown in FIG. 24 and FIG. 25, for the Y-branch waveguide using the rectangular waveguide described in Non-Patent Document 1, a graph obtained by simulating the change in excess loss with respect to wavelength is shown. 31 and FIG. FIG. 31 is a graph showing excess loss at the time of conversion from the even mode to the TE ₀ mode, and FIG. 32 is a graph showing excess loss at the time of conversion from the odd mode to the TE ₁ mode. Note that the way of viewing the graphs shown in FIGS. 31 and 32 is the same as that shown in FIGS.

From the graphs shown in FIGS. 31 and 32, the change in excess loss during conversion from the even mode to the TE ₀ mode was 0.61 dB in the wavelength range of 1480 to 1680 nm. When the waveguide width fluctuated by ± 30 nm, the change in excess loss during conversion from the even mode to the TE ₀ mode was 0.19 dB at maximum in the wavelength range of 1480 to 1680 nm.

From this result, it can be seen that Example 1 is more resistant to manufacturing errors than Comparative Example 1. This is due to the above effect 2.
Therefore, it was revealed that the Y branch waveguide of Example 1 is superior to the Y branch waveguide of Comparative Example 1.

(Example 2)
Example 2 is an example corresponding to the substrate-type optical waveguide device 101 of the second embodiment, and the dimensions of each part in this Y-branch waveguide are shown in FIGS. 33 (A), (B), (C). As shown in FIG. 33A is a plan view showing the dimensions of each part. FIG. 33B is a cross-sectional view taken along line Z ₅ -Z ₅ in FIG. FIG. 33C is a cross-sectional view taken along line Z ₆ -Z ₆ in FIG. That is, the dimension of each part of the Y branch waveguide in the second embodiment is the same as the dimension of each part of the Y branch waveguide in the first embodiment. The excess loss is calculated using the butt coupling portion 31.

  Similarly to the case shown in FIG. 23, FIG. 34 shows a graph obtained by simulating the change in excess loss with respect to the width X [nm] of the output waveguide for the Y branch waveguide of Example 2. Note that the way of viewing the graph shown in FIG. 34 is the same as that shown in FIG.

From the graph shown in FIG. 34, the change in excess loss with respect to the width X [nm] of the output waveguide has the same tendency as the graph shown in FIG. Specifically, when X = about 1900 nm, the excess loss of the TE ₀ mode takes the minimum value, and the excess loss at this time is 0.32 dB in the conversion from the even mode to the TE ₀ mode, and from the odd mode to the TE _1. Conversion to mode resulted in 0.49 dB.

  From this result, it can be seen that the Y-branch waveguide of Example 2 can be converted with lower loss loss than the Y-branch waveguide of Comparative Example 1. On the other hand, it can be seen that the loss is larger than that of the Y-branch waveguide of Example 1.

DESCRIPTION OF SYMBOLS 1,101A, 101B ... Substrate type optical waveguide element 2 ... Substrate 3 ... Y branching waveguide 4 ... Core 5 ... Cladding 6 ... First waveguide 7 ... Second waveguide 8 ... Third waveguide 9 ... First 1 rib part 10 ... 2nd rib part 11, 11A, 11B ... 3rd rib part 11a, 11b ... protrusion part 12, 12A, 12B ... slab part 12C ... 1st slab part 12D ... 2nd slab part DESCRIPTION OF SYMBOLS 13 ... Lower clad 14 ... Upper clad 15 ... 1st bending waveguide 16 ... 2nd bending waveguide 17 ... 4th waveguide 18 ... Linear waveguide 31 ... Butt coupling part (branch waveguide structure) 32, 33 ... Input waveguide 34 ... Output waveguide 102,103 ... Tapered waveguide 50A, 50B, 70A, 70B ... Mach-Zehnder optical modulator 51 ... Input part 51A ... First input part 51B ... Second input part 52 ... minute Wave part 52A ... 1st Demultiplexing unit 52B ... second demultiplexing unit 53,54 ... phase modulation unit 55 ... multiplexing unit 55A ... first multiplexing unit 55B ... second multiplexing unit 56 ... output unit 56A ... first output unit 56B ... Second output unit 57 ... High-order mode splitter 58 ... Photo detector 59 ... First light modulation unit 60 ... Second light modulation unit 61 ... Phase adjustment unit 71 ... High-order polarization conversion unit 72 ... Bias Wave beam splitters 80A, 80B ... Mach-Zehnder type output polarization conversion element 81 ... Input unit 82 ... Demultiplexing unit 83 ... Multiplexing unit 84 ... Output unit 85 ... Mach-Zehnder interferometer 86 ... High-order polarization conversion unit 87 ... Phase adjustment unit 88, 89 ... Phase modulation unit 90 ... Optical modulation unit

Claims (13)

On a substrate, a substrate-type optical waveguide device comprising a core and a clad that covers the core and has a refractive index smaller than that of the core,
A first waveguide and a second waveguide in parallel in parallel with each other, and a third waveguide whose other end is connected to one end of said first waveguide and said second waveguide, the third A branch waveguide structure having a fourth waveguide with the other end connected to one end of the waveguide,
The core includes a first rib portion having a rectangular cross section forming the first waveguide, a second rib portion having a rectangular cross section forming the second waveguide, and the third waveguide. And a third rib portion having a rectangular cross section forming the fourth waveguide , and a height lower than the first rib portion, the second rib portion, and the third rib portion. A slab portion shared between one rib portion, the second rib portion, and the third rib portion,
In the first waveguide and the second waveguide, the widths of the first rib portion and the second rib portion are constant in the respective length directions, and the first rib portion and the second waveguide portion are Forming a linear waveguide in which the distance between the second rib portion is constant in the length direction of the first rib portion and the second rib portion;
In the third waveguide, the center in the width direction of the third rib portion coincides with the center in the width direction between the first rib portion and the second rib portion, and the third waveguide The width of the third rib portion is greater than the sum of the width of the first rib portion, the width of the second rib portion, and the distance between the first rib portion and the second rib portion. Forming a large linear waveguide ,
The fourth waveguide forms a tapered waveguide in which the width of the third rib portion gradually decreases toward the side opposite to the side connected to the third waveguide;
A portion of the third rib portion where the other end of the fourth waveguide is formed has the same width and thickness as the portion where the third waveguide is formed. Type optical waveguide element.   2. The substrate-type optical waveguide device according to claim 1, wherein the slab part is provided continuously on both sides in the width direction of the first rib part and the second rib part.   3. The substrate type optical waveguide device according to claim 1, wherein the slab portion is provided continuously on both sides in the width direction of the third rib portion. 4. A bending waveguide having one end connected to the other end of either or both of the first waveguide and the second waveguide;
The bending waveguide is formed by bending either one or both of the first rib portion and the second rib portion in a plane, so that the first rib portion 4. The substrate-type optical waveguide device according to claim 1, wherein the substrate-type optical waveguide device has a shape in which a distance between the second rib portion is continuously reduced. A tapered waveguide having one end connected to the other end of the bending waveguide;
The taper waveguide is continuously provided on one side or both sides in the width direction of the first rib portion, and continuously on one side or both sides in the width direction of the second rib portion. A second slab portion provided,
The first slab part and the second slab part are provided continuously to the slab part, and each width has a shape that continuously increases toward the slab part. The substrate-type optical waveguide device according to claim 4. It said core comprises Si, planar optical waveguide device according to any one of claim 1 to 5, wherein the clad is characterized in that it comprises a SiO _2. An input unit to which light is input;
A demultiplexing unit for demultiplexing the light input from the input unit;
At least one phase modulating unit for phase modulating the light demultiplexed by the demultiplexing unit;
A multiplexing unit for multiplexing the light phase-modulated by the phase modulation unit;
An output unit that outputs the light combined by the combining unit,
An optical modulator using the substrate type optical waveguide device according to any one of claims 1 to 6 for the multiplexing unit. A first input unit to which light is input, a first demultiplexing unit for demultiplexing light input from the first input unit, and a phase of the light demultiplexed by the first demultiplexing unit Output at least one phase modulation unit to be modulated, a first multiplexing unit for multiplexing the light modulated by the phase modulation unit, and light combined by the first multiplexing unit A first light modulation unit and a second light modulation unit each having a first output unit;
A second input unit that is positioned before the first light modulation unit and the second light modulation unit and receives light input from the second input unit, and the first input unit receives light from the second input unit. A second demultiplexing unit that demultiplexes into a first input unit side of the light modulation unit and a first input unit side of the second light modulation unit;
The light output from the first output unit of the first light modulation unit, located after the first light modulation unit and the second light modulation unit, and the second light modulation unit A second combining unit that combines the light output from the first output unit, and a second output unit that outputs the light combined by the second combining unit,
An optical modulator comprising the substrate-type optical waveguide device according to any one of claims 1 to 6 , wherein the substrate-type optical waveguide element is used in any of the first multiplexing unit and the second multiplexing unit. A phase adjustment unit that adjusts a phase difference of light output from the first multiplexing unit of any one of the first light modulation unit and the second light modulation unit. Item 9. The optical modulator according to Item 8 . 10. The optical modulator according to claim 7 , wherein, in the TE mode in which the electric field is in the in-plane direction of the substrate, the TE ₀ mode light representing a mode having the largest effective refractive index; The TE ₁ mode light representing the mode with the second highest effective refractive index is a monitor structure of an optical modulator that separates and detects the TE ₁ mode light when it is simultaneously output from the output unit. And
A higher-order mode splitter, which is located at a subsequent stage of the output unit and separates the light of the TE ₁ mode;
And a photodetector for detecting the TE ₁ mode light separated by the high-order mode splitter. 10. The optical modulator according to claim 7 , wherein, in the TE mode in which the electric field is in the in-plane direction of the substrate, the TE ₀ mode light representing a mode having the largest effective refractive index; , the TE ₁ mode effective refractive index represents the high mode to the second and light, when it is simultaneously output from the output unit, the light of the TE ₁ mode, the electric field of the TM mode which is perpendicular direction of the substrate A light modulator monitor structure for separating and detecting the TE ₀ mode light after converting it to a T M ₀ mode light representing a mode having the largest effective refractive index in the medium;
A higher-order polarization converter that converts the TE _1- mode light into the T M _0- mode light;
A polarization beam splitter for separating the TE ₀ mode light;
And a photodetector for detecting the TE ₀ mode light separated by the polarization beam splitter. An input unit to which light is input; a demultiplexing unit for demultiplexing the light input from the input unit; a multiplexing unit for multiplexing the light demultiplexed by the demultiplexing unit; and the multiplexing unit An Mach-Zehnder interferometer having an output unit for outputting the combined light;
The effective refractive effective index in the TE mode electric field is in-plane direction of the substrate is the light of TE ₁ mode indicating a high mode second, in the T M mode in the electric field are perpendicular direction of the substrate A high-order polarization converter that converts light into a T M ₀ mode representing the mode with the highest rate;
A phase adjustment unit that adjusts a phase difference of any one of the lights demultiplexed by the demultiplexing unit,
The output polarization conversion element characterized by using the substrate type optical waveguide element according to any one of claims 1 to 6 for the multiplexing part. 13. The output polarization conversion element according to claim 12 , further comprising two phase modulation units that phase-modulate the light demultiplexed by the demultiplexing unit.